Pulmonary infections are a major cause of morbidity and mortality in patients with hematologic malignancy. Bronchoscopy is at present still the traditional first investigation in immunosuppressed patients that have developed pulmonary infiltrates. There is limited data available on the validity of fiberoptic bronchoscopy (FOB) with bronchoalveolar lavage (BAL) to determine the etiology of pulmonary infiltrates with concurrent hematologic malignancy. We retrospectively analyzed the microbiological results of 206 bronchoscopic examinations and treatment changes used in 187 patients with hematologic malignancy and pulmonary infiltrates. Bacteria, fungi, and viruses were found in 85 (41.3 %), 49 (23.8 %), and 55 (28.6 %) of cases, respectively, and overall yield of bronchoscopy was 65.0 %. We compared the microbiological findings with respect to neutropenia, hematopoietic stem cell transplantation (HSCT) status, and the type of malignancy. There were significantly more bacterial and viral infections detected in post-HSCT patients, and more viruses were detected in patients without neutropenia. Galactomannan (GM) was measured in 149 BAL samples. With a GM index threshold of ≥0.5, the sensitivity, specificity, and positive and negative predictive values (PPV and NPV, respectively) of the BAL GM assay were 93.94 %, 86.21 %, 65.96 %, and 98.04 %, respectively. Treatment was modified in 62 cases (30.1 %). There was no significant relationship of treatment modification with the underlying disease, HSCT, or neutropenia. Bronchoscopy with BAL is a valuable diagnostic tool to determine the etiology and appropriate treatment in patients with hematologic malignancy and pulmonary infiltrates. A BAL GM test is recommended when invasive pulmonary aspergillosis is suspected.